The Calculation of Spatial Prices in the Absence of Unit Values: Alternative Methodologies with Empirical Evidence from India

The calculation of spatial price indices in a country commonly referred to as ‘Regional Purchasing Power Parity’ (RPP), is important in case of large countries with heterogeneous preferences and prices between regions. While spatial price differences between countries have featured prominently in the International Comparison Project (ICP) of the United Nations that periodically calculates the ‘Purchasing Power Parity’ (PPP) of a country’s currency, spatial price differences within a country has received much less attention. This is largely due to the absence on price information on the same group of items from different regions in the country. This paper proposes alternative procedures for estimating spatial prices that either require no spatial price information or require only limited information in the form of temporal price indices at an aggregated level of items that varies between regions. The alternative procedures differ in that while one is based on the idea of a ‘true cost of living index’ (TCLI) that is based on consumer preferences and requires demand estimation, the other is based on the Barten (1964) notion that demographic variation between households can be viewed as ‘quasi-price’. The former has the advantage in not requiring the hypothesis of ‘quasi-price’ demographic effects that has been rejected by Muellbauer (1977) on UK data. The present evidence on spatial price differences in India, which is the most comprehensive to date since it goes down to district kevel, shows that both the proposed procedures have considerable potential in future applications on other data sets with limited price information. Since the procedures allowed the study to be conducted on the full basket of items, and not just the food items, the RPPs presented in this paper are more useful than the earlier RPPs for India. The sensitivity of the demand elasticities and expenditure inequalities to the use of spatial price deflators establishes the importance of the RPPs in policy applications. For example, the TCLI based spatial price deflators of nominal household expenditures show that the omission of spatial prices will lead to an overstatement of expenditure inequality since the more affluent states/districts are also the ones with higher cost of living.